Impact of baseline left atrial function on long-term outcome after catheter ablation for paroxysmal atrial fibrillation

Background: Left atrial (LA) size is an established predictor of recurrence after catheter ablation for paroxysmal atrial fibrillation (PAF). We investigated the impact of baseline LA function on recurrence after PAF ablation and compared the predictive values of LA function with those of LA size. Methods: We retrospectively investigated 292 consecutive patients who underwent PAF ablation (median follow-up: 3.0 years). All patients had their preoperative LA volume (LAV) assessed using cardiac computed tomography under sinus rhythm. We used LA emptying fraction (LAEF) as an indicator of LA function and assessed the association between baseline LAEF and recurrence after initial ablation using a multivariate Cox hazard model. Then, we performed receiver operating characteristic analysis for predicting recurrence after single and multiple procedures and compared the c-statistics of LAEF and indexed maximum and minimum LAV (LAVI(max) and LAVI(min)) Results: In a multivariate Cox hazard model, LAEF was strongly associated with recurrence after a single procedure [hazard ratio (HR): 0.968, 95% confidence interval (CI): 0.951-0.985, p < 0.001]. In the receiver operating characteristic analysis for predicting recurrence, the predictive accuracy of LAEF was mild after a single procedure [area under the curve (AUC): 0.666, p < 0.001] and moderate after multiple procedures (AUC: 0.701, p < 0.001). The c-statistic of LAEF was significantly higher than those of LAVI(max) and LAVI(min) after a single procedure (p < 0.05, for both). After adjustment for factors related to reduced LAEF (increased serum brain natriuretic peptide, age, LA diameter, and reduced left ventricular ejection fraction), it was still associated with recurrence (HR: 0.964, 95% CI: 0.946-0.982, p < 0.001). Conclusions: LAEF was associated with recurrence after PAF ablation. LA function is a more useful predictor than LA size. (C) 2019 Japanese College of Cardiology. Published by Elsevier Ltd. All rights reserved.